TAM A320 crash at Congonhas, Brazil
Join Date: Aug 2002
Location: Switzerland, Singapore
Posts: 1,309
Received 0 Likes
on
0 Posts
BOAC, thanks, I was wrong, you and GMDS are right. Still it is no argument for *his* argument since the same would haven to any other conventional aircraft with autothrottle: If you reduce only one TL and let the other keep the existing speed. Otherwise you couldn't survive any engine failure!
To my knowledge I assume that the EPR1 value is for full reverse.
Bernd, Alpha Floor protection is nothing Airbus has alone. Most airliners with autothrottle have it, too. It is also no argument for a non moving trust lever.
Dani
To my knowledge I assume that the EPR1 value is for full reverse.
Bernd, Alpha Floor protection is nothing Airbus has alone. Most airliners with autothrottle have it, too. It is also no argument for a non moving trust lever.
Dani
Join Date: Jul 2007
Location: Germany
Posts: 556
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by BOAC
Dani - re-read the FDR please. No 2 thrust increased (as an A/T function to try to maintain set speed we are told) out,
until the A/T dropped when it decreased again to level at (presumably) the EPR for the EITHER the T/L position or, if there had been some sort of sensor failure, the EPR for where the system THOUGHT the T/L was.
The decrease in the graphs is more likely due to the fact that an increase in speed was achieved by the increased thrust, and so the feedback-control loop reduced thrust again.
During approach autothrust speed-mode, control is more complicated than during cruise flight, and takes into account both air- and ground speed. In the graphs we can see that the speed drops a bit after touchdown (18:48:24) and thrust increases, then speed increases slightly (18:48:32), after which thrust drops again, and then A/THR disconnects (18:48:29.5), freezing thrust.
Recorded thrust lever angle was CL detent (22 degrees or so), thrust associated with that would have been close to EPR 1.28 (see FDR graphs page 11, showing a take-off until shortly after thrust reduction.)
Thrust only changes to meet actual thrust lever position when A/THR is disconnected with the thumb-push-button on the side of the thrust levers (the standard way).
I'm still waiting for an answer to whether the EPR on the reversed No 1 represents full reverse or idle?
Originally Posted by GMDS
It might even be construed in such a way that you can choose between fixed TL, moving TL and manual TL.
Bernd
Join Date: Jul 2007
Location: Germany
Posts: 556
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by Dani
Bernd, Alpha Floor protection is nothing Airbus has alone. Most airliners with autothrottle have it, too. It is also no argument for a non moving trust lever.
Whether or not it is an argument really relies on whether TOGA-thrust at Alpha-Floor-Protection also moves the levers fully forward, and if so, on crew reaction to that.
(I can imagine that this sort of emergency function does not move the levers even on moving levers-systems; on Airbus, too, Alpha-Floor-Protection is abnormal in that it is the only case in which autothrust can be active outside its normal active range of thrust lever settings, and also the only situation in which thrust can be higher than lever setting.)
And I will also say that, as to my 2nd argument, if moving thrust levers could be shown to have a significant safety advantage (which has not been shown), then this convenience argument is less important.
Bernd
Join Date: Jul 2007
Location: the City by the Bay
Posts: 547
Likes: 0
Received 0 Likes
on
0 Posts
Being very simplistic here, but perhaps Airbus should rethink their aircraft control logic and follow more the KISS principle in their basic design concept. Making their planes even easier to fly and more intuitive to the human brain . And making sure that those that pilot them are more then adequately trained.
Airbus has had more man/machine problems then the competition throughout the last few decades it would appear. Airbus aircraft have the final say , rather then the pilots in any pilot / plane confrontation. Unless you knew what the plane was designed to do and never ever forgot any part of it, you ARE in danger, because YOU wont be able to dictate what it does and wont be able to understand what it is doing.
Airbus has had more man/machine problems then the competition throughout the last few decades it would appear. Airbus aircraft have the final say , rather then the pilots in any pilot / plane confrontation. Unless you knew what the plane was designed to do and never ever forgot any part of it, you ARE in danger, because YOU wont be able to dictate what it does and wont be able to understand what it is doing.
Join Date: Feb 2007
Location: i don't know
Posts: 320
Likes: 0
Received 0 Likes
on
0 Posts
Bernd
And just who would show that? If humble user-pilots try to reason as to why it might be so, they get a severe bashing. Furthermore they get their legitimacy questioned, because they are ... humble pilots only!! The manufacturor, authorities and airlines, all of them self proclaimed legitimate to "prove" or "disapprove" only, have no interest whatsoever to go ahead on this because it would: 1)cost, 2)show some of their ideas were wrong), 3)incite lawsuits, 4)sell less airplanes, 5)mean less "contributions" aso. A lost cause.
By the way: Has it been shown that a fixed throttle has a significant safety advantage?
Your argument about Alpha floor only withstands with the assumption that the pilot would interfere. Just as yours, Dani, that many pilots "fight" the moving AT.
I have to agree. Just as the idea of relying on HMI designers.
A "choice" for his subjects has always been a nightmare for any ruler .... (sarcasm out)
GMDS
And I will also say that, as to my 2nd argument, if moving thrust levers could be shown to have a significant safety advantage (which has not been shown), then this convenience argument is less important.
By the way: Has it been shown that a fixed throttle has a significant safety advantage?
Your argument about Alpha floor only withstands with the assumption that the pilot would interfere. Just as yours, Dani, that many pilots "fight" the moving AT.
Quote:
Originally Posted by GMDS
It might even be construed in such a way that you can choose between fixed TL, moving TL and manual TL.
Terrible idea, as every HMI designer will tell you.
Originally Posted by GMDS
It might even be construed in such a way that you can choose between fixed TL, moving TL and manual TL.
Terrible idea, as every HMI designer will tell you.
A "choice" for his subjects has always been a nightmare for any ruler .... (sarcasm out)
GMDS
Paxing All Over The World
Non-pilot speaking
A question I have long been waiting to be answered in this thread and, as it has not, now ask the question:
What is the operating procedure when a reverse is inop?
Is it better to be able to: close both TL's; select REV and move both (or 'all' if more than 2) TLs and not have to worry that one of the REVs is inop?
A question I have long been waiting to be answered in this thread and, as it has not, now ask the question:
What is the operating procedure when a reverse is inop?
- The equipment is physically isolated (by electrical, hydraulic or mechanical means) to that, even if you select REV, it will not happen but the engine will spool up as the TL is moved and provide forward thrust.
- The equipment is Isolated (as above) but you are instructed NOT to select REV and NOT move the TL from idle, after closing thrust.
- Some more sophisticated isolation that might involve ensuring that, even if you select REV and move the TL from idle - the engine will remain at idle and the REV equipment will not deploy. That is, some form of electronic - rather than physical - isolation is in place.
Is it better to be able to: close both TL's; select REV and move both (or 'all' if more than 2) TLs and not have to worry that one of the REVs is inop?
Join Date: Aug 2007
Location: Brazil
Age: 71
Posts: 131
Likes: 0
Received 0 Likes
on
0 Posts
I understand we have Airbus and Boeing pilots here.
Quick question: If a TL was to be "forgotten" during flare (one reverse inop), in which TL system this would be more likely to happen - moving or non moving TLs?
Quick question: If a TL was to be "forgotten" during flare (one reverse inop), in which TL system this would be more likely to happen - moving or non moving TLs?
Sun worshipper
Join Date: Dec 2001
Location: Paris
Posts: 494
Likes: 0
Received 0 Likes
on
0 Posts
GMDS,
Your argument about Alpha floor only withstands with the assumption that the pilot would interfere. Just as yours, Dani, that many pilots "fight" the moving AT.
The Air China A300 in Nagoya was in the same type of scenario. Only far worse.
A question that bugs me is why people do refuse to consider the human factor (s) in this type of accident and only concentrate on the M /non M argument ?
Dani,
Bernd, Alpha Floor protection is nothing Airbus has alone. Most airliners with autothrottle have it, too. It is also no argument for a non moving trust lever.
Regards.
Join Date: Jul 2007
Location: the City by the Bay
Posts: 547
Likes: 0
Received 0 Likes
on
0 Posts
Yes the China Airlines A300-600 disaster in Nagoya is a very good example of the pilots unwittingly and unknowingly fighting with the aircraft's designed operating principles. And I am not clear why the power failed after the aircraft stalled, after which it fell to the ground.
That particular incident is worth VERY careful study.
http://www.rvs.uni-bielefeld.de/publ...ei/taipei.html
http://www.rvs.uni-bielefeld.de/publ...goya_a300.html
That particular incident is worth VERY careful study.
http://www.rvs.uni-bielefeld.de/publ...ei/taipei.html
http://www.rvs.uni-bielefeld.de/publ...goya_a300.html
Join Date: Aug 2007
Location: The Netherlands
Age: 67
Posts: 288
Likes: 0
Received 0 Likes
on
0 Posts
Rate
Bernd,
This accident at a rate of 1 per 17 million landings - no, the rate should be calculated versus the number of landings with one thrust reverser deactivated and the pilots using autothrust for the approach.
Rate would be closer to 1 per 17.000 landings then.
This accident at a rate of 1 per 17 million landings - no, the rate should be calculated versus the number of landings with one thrust reverser deactivated and the pilots using autothrust for the approach.
Rate would be closer to 1 per 17.000 landings then.
Guest
Posts: n/a
Members of the Brazilian Chamber of Deputies had access today to the actual voice recording on the last 30 minutes of the PR-MBK.
I can't figure out what made one of the Congressmen say that after listening to the recordings and meeting with CENIPA personnel he got definitively convinced that the accident was not caused by HF.
I can't figure out what made one of the Congressmen say that after listening to the recordings and meeting with CENIPA personnel he got definitively convinced that the accident was not caused by HF.
Join Date: Aug 2002
Location: Switzerland, Singapore
Posts: 1,309
Received 0 Likes
on
0 Posts
Paxboy, on an Airbus you can never have trust increase on an engine with actual mechanical failure of a TR. The system is designed that if there's something wrong, there is no trust added. But in the actual case this was different because the TL remained in CLB so the pilots deliberately kept trust. Older aircraft where very different, as you might remember the latest NWA accident, where a MD-80 hit a A320 on ground with high speed (hydraulics failure, no reverser buckets, but full "revers" trust). .
In fact, in most cases there would be no harm if you would go to full revers even if the TR are u/s. Please check also the FDR from the previous landing, where the RHS pilot went into revers on both engines, and nothing happened. Airbus recommends not to go into revers if one is u/s, to cover all technical failures.
Studi, you have to leave the authority to the pilots. If you loose one engine during flare, or if you should get a reverser deployed in the air, you cannot automatically retard the other. Airbus is inspite of the advanced automatics always giving the last authority to the pilots. If they want trust, they get it. If you want trust, you leave the TL in CLB. A computer cannot read a human's mind! You have to change the mind of these pilots, not the automatics!
Lemurian, ANY airliner that I know add trust before they stall. Fokker 100 come to my mind, MD-11... I'm surprised that Boeing still doesn't do it. Maybe in the 787 finally... I'm so glad I'm not flying these oldtimers
Dani
In fact, in most cases there would be no harm if you would go to full revers even if the TR are u/s. Please check also the FDR from the previous landing, where the RHS pilot went into revers on both engines, and nothing happened. Airbus recommends not to go into revers if one is u/s, to cover all technical failures.
Studi, you have to leave the authority to the pilots. If you loose one engine during flare, or if you should get a reverser deployed in the air, you cannot automatically retard the other. Airbus is inspite of the advanced automatics always giving the last authority to the pilots. If they want trust, they get it. If you want trust, you leave the TL in CLB. A computer cannot read a human's mind! You have to change the mind of these pilots, not the automatics!
Lemurian, ANY airliner that I know add trust before they stall. Fokker 100 come to my mind, MD-11... I'm surprised that Boeing still doesn't do it. Maybe in the 787 finally... I'm so glad I'm not flying these oldtimers
Dani
Last edited by Dani; 5th Sep 2007 at 07:46.
Join Date: Aug 2002
Location: Switzerland, Singapore
Posts: 1,309
Received 0 Likes
on
0 Posts
Studi, I agree that these are situations where you shouldn't be in. In eighter case!
But sometimes you are! If it's your bad day, you have assymetrical trust, you have one reverser working and the other not.
1st case: you want to stop but do not have idle yet.
That's exactly what the system did! One TL was not in idle. Hence no braking. As long as a pilot wishes to have more than idle, the aircraft should not be brought to a stand still. The system design had completly anticipated the will of the pilot. Unfortunately that wasn't what the pilots wanted. Once again. A machine cannot read one's mind.
2nd case: you want to continue but have reverser
Yes there is. Unintended reverser deployment. Mishandling of TL. Human error. Mechanical failure...
The system also anticipates this situation by adding the trust on the remaining engine.
Both cases are built in the system. It is fool proof. Believe me, those engineers spent a lot of time to go through all posibilities.
I don't know if Rananim is right about what he says about Boeing, but also if a Boeing doesn't give you reverse if you are not in idle, this doesn't help you in the TAM case. They never wanted to touch and go, never go around. They wanted to land, but "forgot" the TL. They could have never gone around, when they realized they were in big trouble.
There are a lot of people with doubts about AI technology here in this forum. I understand that, since they have never seen the system from close. Ignorance is the root of prejudice. But any aircraft in the world would have behaved in this TAM case like the A320. This is not an Airbus case, it is a HF case. The longer we ignore it, the longer we don't find the reason for the tragedy.
Dani
But sometimes you are! If it's your bad day, you have assymetrical trust, you have one reverser working and the other not.
1st case: you want to stop but do not have idle yet.
Why not block reverse until both levers are in IDLE
2nd case: you want to continue but have reverser
There is absolutely never ever the need for reverse on one engine and forward thrust on the other
The system also anticipates this situation by adding the trust on the remaining engine.
Both cases are built in the system. It is fool proof. Believe me, those engineers spent a lot of time to go through all posibilities.
I don't know if Rananim is right about what he says about Boeing, but also if a Boeing doesn't give you reverse if you are not in idle, this doesn't help you in the TAM case. They never wanted to touch and go, never go around. They wanted to land, but "forgot" the TL. They could have never gone around, when they realized they were in big trouble.
There are a lot of people with doubts about AI technology here in this forum. I understand that, since they have never seen the system from close. Ignorance is the root of prejudice. But any aircraft in the world would have behaved in this TAM case like the A320. This is not an Airbus case, it is a HF case. The longer we ignore it, the longer we don't find the reason for the tragedy.
Dani
Join Date: Aug 2007
Location: The Netherlands
Age: 67
Posts: 288
Likes: 0
Received 0 Likes
on
0 Posts
Good old Boeing...
Dani,
Already B-767/757 autothrottles have Alpha mode, which will not allow speed to go lower, if limit Angle of Attack is reached. This implies that TOGA thrust will be given, if that is what it takes to prevent further speed decay.
Just off the top of my head, difference between AI and B setup may be in the fact whether A/T has to be engaged or not, for this protection to work.
-------------------------------------------------
More or less the same goes for A/P modes - they will also respect speed limits, but do so when engaged. In B, if pilot manually puts the a/c on it's nose or back, the flight controls won't stop him. In A (FBW) a/c, the pilot is prevented, also in manual flight, from performing outlandish deeds by the flight control system.
-------------------------------------------------
This TOGA lock thing gives me another argument to illustrate the difference between moving and non-moving thrust levers.
To get out of TOGA lock in the A-320, you have to disengage A/T.
If you just pull back the thrust levers, nothing will happen. You would notice that nothing would happen, for instance, by looking at the engine instruments. Then you would have to consciously reason why the engine indications have not reduced to idle and then you would realize - oh, got to disconnect A/T.
In moving thrust lever setup, you may also have to disengage A/T to cancel ALPHA mode, but if you just pull levers to idle without disconnecting, the engines will go to idle and your hand will feel the A/T working against your hand, trying to move the levers forward again. In this way, you get what you want immediately, not only after the observation, reasoning, action cycle has been completed, and, the road towards realizing what you have forgotten may be more direct - from the hand that feels opposition directly to realization, without a knowledge based reasoning process.
EMIT.
Already B-767/757 autothrottles have Alpha mode, which will not allow speed to go lower, if limit Angle of Attack is reached. This implies that TOGA thrust will be given, if that is what it takes to prevent further speed decay.
Just off the top of my head, difference between AI and B setup may be in the fact whether A/T has to be engaged or not, for this protection to work.
-------------------------------------------------
More or less the same goes for A/P modes - they will also respect speed limits, but do so when engaged. In B, if pilot manually puts the a/c on it's nose or back, the flight controls won't stop him. In A (FBW) a/c, the pilot is prevented, also in manual flight, from performing outlandish deeds by the flight control system.
-------------------------------------------------
This TOGA lock thing gives me another argument to illustrate the difference between moving and non-moving thrust levers.
To get out of TOGA lock in the A-320, you have to disengage A/T.
If you just pull back the thrust levers, nothing will happen. You would notice that nothing would happen, for instance, by looking at the engine instruments. Then you would have to consciously reason why the engine indications have not reduced to idle and then you would realize - oh, got to disconnect A/T.
In moving thrust lever setup, you may also have to disengage A/T to cancel ALPHA mode, but if you just pull levers to idle without disconnecting, the engines will go to idle and your hand will feel the A/T working against your hand, trying to move the levers forward again. In this way, you get what you want immediately, not only after the observation, reasoning, action cycle has been completed, and, the road towards realizing what you have forgotten may be more direct - from the hand that feels opposition directly to realization, without a knowledge based reasoning process.
EMIT.
Join Date: Jul 2007
Location: Germany
Posts: 556
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by Dani
Lemurian, ANY airliner that I know add trust before they stall. Fokker 100 come to my mind, MD-11... I'm surprised that Boeing still doesn't do it. Maybe in the 787 finally... I'm so glad I'm not flying these oldtimers
This is not the same as Airbus's "Alpha-Floor-Protection" on their FBW airliners.
I read up a bit on the FBW Boeing (there is a thread on PPrune under the Tech Logs about Flight Envelope Protection).
While that machine makes it difficult, with sufficient force on the yoke you can fly it outside the envelope and stall it.
An A320 (assuming Normal Law), will not let the pilot exceed a certain angle of attack ("Alpha-Max"), no matter how hard he pulls the stick back.
Before reaching Alpha-Max, (at "Alpha-Floor"), TOGA-thrust is set and locked, and has to be turned off explicitly.
Bernd
Last edited by bsieker; 5th Sep 2007 at 08:46. Reason: Correctly attributed quote.
Join Date: Sep 2000
Location: Bielefeld, Germany
Posts: 955
Likes: 0
Received 0 Likes
on
0 Posts
Originally Posted by studi
There is absolutely never ever the need for [XYZ] Never. So why give the pilots the possibility to do so?
Let me try a short precis of how such systems need to be designed.
First, there are sensors which sense *some* of the system state (including *some* of the intentions of the operators). Then, on the basis of the sensor readings a designer has to decide
(a): what aircraft configuration makes most sense, and
(b) what the hazards are from hisher choice in (a).
Those hazards would be situations in which the aircraft configuration chosen in (a) mismatched the situation the aircraft really was in. And for all the hazards you can think of under (b), you then have to show:
(c) how the hazards can be mitigated.
You want to say there is never a situation in which one wants reverse on one engine and thrust on the other. That may be, but then a designer who is explicitly considering this undesirable situation would need to determine what is best to do. There seem to me two possibilities:
(i) the pilot wants to stop;
(ii) the pilot wants to go.
The hazards are equally obvious:
(x) if you inhibit thrust, the (most obvious) hazard is a situation in which the pilot wants to go;
(y) if you inhibit reverse, the (most obvious) hazard is a situation in which the pilot wants to stop.
How would one design a configuration to admit both these possibilities while mitigating the hazards? Answer: the designer must let the pilot choose in the given situation. Heshe can't choose in advance from hisher comfy desk.
PBL
Last edited by PBL; 5th Sep 2007 at 08:41. Reason: There's always something else ......
Join Date: Aug 2007
Location: The Netherlands
Age: 67
Posts: 288
Likes: 0
Received 0 Likes
on
0 Posts
Thrust levers
To STUDI.
People have repeatedly stated in this forum that B will not allow you to select reverse when *not all* thrust levers are at idle.
They have "proven" this with a quote from FCOM.
What those people have forgotten is, to quote from FCOM a few paragraphs earlier, where it says that *each engine has it's individual control on the flight deck* .
Now as far as I know, individual means - not in any way depending on the other one. Reading and understanding is a complicated activity!!
The thing is, finding out what is really meant in the books, would require playing around with those controls in a live aircraft. Now that is something that you don't do in practice, because, for instance you don't want a ground engineer to get wounded by your fiddling around during a ground stop.
Moreover, why would you ever want to know, if you just do as always, that is, pull alll thrust levers to idle when touching down?
People have repeatedly stated in this forum that B will not allow you to select reverse when *not all* thrust levers are at idle.
They have "proven" this with a quote from FCOM.
What those people have forgotten is, to quote from FCOM a few paragraphs earlier, where it says that *each engine has it's individual control on the flight deck* .
Now as far as I know, individual means - not in any way depending on the other one. Reading and understanding is a complicated activity!!
The thing is, finding out what is really meant in the books, would require playing around with those controls in a live aircraft. Now that is something that you don't do in practice, because, for instance you don't want a ground engineer to get wounded by your fiddling around during a ground stop.
Moreover, why would you ever want to know, if you just do as always, that is, pull alll thrust levers to idle when touching down?